Organization of the file system
? Rename. The possibility of renaming the file creates additional user-friendliness.
This operation can be modeled copying the file to a file with a new name and its
subsequent removal. There are two ways to implement the steps to the files:
In the first case, for each operation performed as universal or unique action (stateless
scheme). For example, the sequence of operations may be this: open, read1, close,…
open, read2, close,… open, read3, close.
An alternative method, which is the universal action performed at the beginning
and the end of the sequence of operations, and for each sub-operations are
performed only unique action. In this case, the above sequence of operations
would look like: open, read1,… read2,… read3, close.
Most OS uses the second method, as a more economical and fast. The first method
is more resistant to failures, because the results of each operation are
independent of the outcome of the previous operation, so it is sometimes used in
distributed systems (such as Sun NFS). Directories. Logical structure file
archive. The number of computer files can be large. Some systems store thousands
of files holding hundreds of gigabytes of disk. Effective management of this
data implies that there are clear logical structure. All modern filesystems
support multilevel naming files from the maintenance of additional external
memory with a special file-directory structure (or directory).
Each directory contains a list of folders and / or files contained in the
directory. Catalogs have the same internal format where each file corresponds to
a record in the file directory. When the system opens the file, it looks for the
file name in the directory. Then extracted attributes of the file and address
blocks on the disk or directly from the entries in the directory or structure to
which the entry points in the directory. This information is placed in a
systemic table in the main memory. All subsequent references to the file are
using this information. The number of directories depends on the system. Earlier
OS had only one root directory, and then there directories for users (one
directory per user). In modern OS used arbitrary directory tree structure. Thus,
the files on the disk form a hierarchical tree-like structure
There are a number of equivalent ways to image the tree inverted tree structure
shown at 11.4, the most popular. The upper vertex called Root. If the tree can
not have offspring, it is called a terminal or a sheet of vertices (in this case
a file). Not flat-tops guides or directories contain lists sheet and no leafy
tops. The path from the root to the file clearly identifies the file. ? Rename.
The possibility of renaming the file creates additional user-friendliness. This
operation can be modeled copying the file to a file with a new name and its
subsequent removal. There are two ways to implement the steps to the files:
In the first case, for each operation performed as universal or unique action (stateless
scheme). For example, the sequence of operations may be this: open, read1, close,…
open, read2, close,… open, read3, close.
An alternative method, which is the universal action performed at the beginning
and the end of the sequence of operations, and for each sub-operations are
performed only unique action. In this case, the above sequence of operations
would look like: open, read1,… read2,… read3, close.
Most OS uses the second method, as a more economical and fast. The first method
is more resistant to failures, because the results of each operation are
independent of the outcome of the previous operation, so it is sometimes used in
distributed systems (such as Sun NFS). Directories. Logical structure file
archive. The number of computer files can be large. Some systems store thousands
of files holding hundreds of gigabytes of disk. Effective management of this
data implies that there are clear logical structure. All modern filesystems
support multilevel naming files from the maintenance of additional external
memory with a special file-directory structure (or directory).
Each directory contains a list of folders and / or files contained in the
directory. Catalogs have the same internal format where each file corresponds to
a record in the file directory. When the system opens the file, it looks for the
file name in the directory. Then extracted attributes of the file and address
blocks on the disk or directly from the entries in the directory or structure to
which the entry points in the directory. This information is placed in a
systemic table in the main memory. All subsequent references to the file are
using this information. The number of directories depends on the system. Earlier
OS had only one root directory, and then there directories for users (one
directory per user). In modern OS used arbitrary directory tree structure. Thus,
the files on the disk form a hierarchical tree-like structure
There are a number of equivalent ways to image the tree inverted tree structure
shown at 11.4, the most popular. The upper vertex called Root. If the tree can
not have offspring, it is called a terminal or a sheet of vertices (in this case
a file). Not flat-tops guides or directories contain lists sheet and no leafy
tops. The path from the root to the file clearly identifies the file.
Next part
|